1. Introduction
Recent tax reforms in most OECD countries have reduced the progressivity of income taxation by lowering marginal tax rates for given average tax rates. The
tax reforms are typically motivated by a desire to increase incentives to work and thereby expand activity in the economy. This view is supported by models with
perfect competition in the labour market: A decrease in the marginal tax rate leads to a substitution effect towards higher labour supply, and as the average tax is
fixed no income effect is present, and so, the effect on activity is positive. We will name this effect the ‘‘labour supply effect’’.
More recent papers have challenged this view by arguing that the assumption of a perfectly competitive labour market is a bad description of most real world
labour markets facing unemployment and therefore also misleading for our understanding of how the structure of the tax system influences the labour market.
In fact, the opposite conclusion is obtained when introducing imperfect competi-
Ž .
Ž .
tion in the labour market, see Hersoug 1984 , Malcomson and Sartor 1987 , Ž
. Ž
. Lockwood and Manning 1993 , and Koskela and Vilmunen 1996 . These papers
consider unionised labour markets where wages are set by unions or negotiated between firms and unions for an exogenously given length of the working day. In
this case, a union faces a trade-off between employment and higher after tax wages for the employed. A decrease in the marginal tax rate of the employed
workers implies that they keep a large fraction of a wage rise for themselves. This increases the incentives of unions to claim higher wages which reduce activity. We
will name this effect the ‘‘union effect’’.
3
One may argue that the two types of results presented above represent polar cases in the sense that the competitive labour market result is a consequence of a
pure labour supply effect whereas the imperfectly competitive result is a conse- quence of no labour supply effect due to the assumption of an exogenous length of
the working day. Therefore, it seems interesting to analyse the case of an imperfectly competitive labour market with an endogenous length of the working
day. We do this by assuming that firms and unions negotiate both wages and number of working hours per worker as is done in many countries, e.g. Denmark.
The result is that the effect of a decrease in the marginal tax rate becomes indeterminate since both a labour supply and a union effect appear. The sign of the
overall effect depends on the bargaining power of the union, the wage elasticity of labour supply, and the progressivity of the tax system. If the initial tax system is
very progressive, it is less likely that the union effect dominates. On the other hand, if the union has a high bargaining power and the elasticity of labour supply
3
This terminology is somewhat misleading as the effect is not specific to a unionized labour market. Ž
. The same effect is found in an efficiency wage setting, see Hoel 1990 , and in search models, see
Ž .
Pissarides 1983 .
is small, then it is more likely that the union effect dominates. The pure union effect found in the aforementioned papers is reproduced in the special case of a
perfectly inelastic labour supply. The theoretical ambiguous effects of changes in tax progressivity motivate
empirical investigation. Therefore, an empirical wage equation is derived from the model and estimated on Danish data for both blue-collar workers and white-collar
workers. The estimations indicate that a reduction in progressivity for blue-collar workers increases real wages and reduces activity, i.e. the union effect dominates,
whereas the estimate for white-collar workers is insignificant maybe reflecting that the union effect and the labour supply effect cancel out.
The remainder of the paper is organised as follows. Section 2 derives the theoretical consequences of changing tax progressivity and the wage equation used
for the estimation. Section 3 estimates the wage equation on both blue-collar workers and white-collar workers and discusses the implications. Section 4
contains concluding remarks.
2. Theory
Kategori